scholarly journals Few-Shot Bayesian Imitation Learning with Logical Program Policies

2020 ◽  
Vol 34 (06) ◽  
pp. 10251-10258
Author(s):  
Tom Silver ◽  
Kelsey R. Allen ◽  
Alex K. Lew ◽  
Leslie Pack Kaelbling ◽  
Josh Tenenbaum
Keyword(s):  
Learning Algorithm ◽  
Training Data ◽  
Policy Learning ◽  
Domain Specific Language ◽  
Computationally Efficient ◽  
Domain Specific ◽  
Strategy Games ◽  
Approximate Bayesian ◽  
Data Requirements ◽  
Approximate Bayesian Inference

Humans can learn many novel tasks from a very small number (1–5) of demonstrations, in stark contrast to the data requirements of nearly tabula rasa deep learning methods. We propose an expressive class of policies, a strong but general prior, and a learning algorithm that, together, can learn interesting policies from very few examples. We represent policies as logical combinations of programs drawn from a domain-specific language (DSL), define a prior over policies with a probabilistic grammar, and derive an approximate Bayesian inference algorithm to learn policies from demonstrations. In experiments, we study six strategy games played on a 2D grid with one shared DSL. After a few demonstrations of each game, the inferred policies generalize to new game instances that differ substantially from the demonstrations. Our policy learning is 20–1,000x more data efficient than convolutional and fully convolutional policy learning and many orders of magnitude more computationally efficient than vanilla program induction. We argue that the proposed method is an apt choice for tasks that have scarce training data and feature significant, structured variation between task instances.

scholarly journals Scalable approximate Bayesian inference for particle tracking data

2018 ◽  
Author(s):  
Ruoxi Sun ◽  
Liam Paninski
Keyword(s):  
Bayesian Inference ◽  
Particle Tracking ◽  
Training Data ◽  
Data Types ◽  
Forward Algorithm ◽  
Bayesian Approaches ◽  
Uncertainty Estimates ◽  
Approximate Bayesian ◽  
Uncertainty Information ◽  
Approximate Bayesian Inference

AbstractMany important datasets in physics, chemistry, and biology consist of noisy sequences of images of multiple moving overlapping particles. In many cases, the observed particles are indistinguishable, leading to unavoidable uncertainty about nearby particles’ identities. Exact Bayesian inference is intractable in this setting, and previous approximate Bayesian methods scale poorly. Non-Bayesian approaches that output a single “best” estimate of the particle tracks (thus discarding important uncertainty information) are therefore dominant in practice. Here we propose a flexible and scalable amortized approach for Bayesian inference on this task. We introduce a novel neural network method to approximate the (intractable) filter-backward-sample-forward algorithm for Bayesian inference in this setting. By varying the simulated training data for the network, we can perform inference on a wide variety of data types. This approach is therefore highly flexible and improves on the state of the art in terms of accuracy; provides uncertainty estimates about the particle locations and identities; and has a test run-time that scales linearly as a function of the data length and number of particles, thus enabling Bayesian inference in arbitrarily large particle tracking datasets.

Scalable Approach to High Coverages on Oxides via Iterative Training of a Machine-Learning Algorithm

2019 ◽  
Author(s):  
Andrew Medford ◽  
Shengchun Yang ◽  
Fuzhu Liu
Keyword(s):  
Machine Learning ◽  
Chemical Potential ◽  
Learning Algorithm ◽  
Absolute Error ◽  
Low Energy ◽  
Training Data ◽  
High Coverage ◽  
Metal Compounds ◽  
Adsorption Energies ◽  
The Stability

Understanding the interaction of multiple types of adsorbate molecules on solid surfaces is crucial to establishing the stability of catalysts under various chemical environments. Computational studies on the high coverage and mixed coverages of reaction intermediates are still challenging, especially for transition-metal compounds. In this work, we present a framework to predict differential adsorption energies and identify low-energy structures under high- and mixed-adsorbate coverages on oxide materials. The approach uses Gaussian process machine-learning models with quantified uncertainty in conjunction with an iterative training algorithm to actively identify the training set. The framework is demonstrated for the mixed adsorption of CH<sub>x</sub>, NH<sub>x</sub> and OH<sub>x</sub> species on the oxygen vacancy and pristine rutile TiO<sub>2</sub>(110) surface sites. The results indicate that the proposed algorithm is highly efficient at identifying the most valuable training data, and is able to predict differential adsorption energies with a mean absolute error of ~0.3 eV based on <25% of the total DFT data. The algorithm is also used to identify 76% of the low-energy structures based on <30% of the total DFT data, enabling construction of surface phase diagrams that account for high and mixed coverage as a function of the chemical potential of C, H, O, and N. Furthermore, the computational scaling indicates the algorithm scales nearly linearly (N<sup>1.12</sup>) as the number of adsorbates increases. This framework can be directly extended to metals, metal oxides, and other materials, providing a practical route toward the investigation of the behavior of catalysts under high-coverage conditions.

scholarly journals Comparing two sequential Monte Carlo samplers for exact and approximate Bayesian inference on biological models

2017 ◽  
Vol 14 (134) ◽  
pp. 20170340 ◽  
Author(s):  
Aidan C. Daly ◽  
Jonathan Cooper ◽  
David J. Gavaghan ◽  
Chris Holmes
Keyword(s):  
Bayesian Inference ◽  
Bayesian Methods ◽  
Sequential Monte Carlo ◽  
Model Parameters ◽  
Biological Models ◽  
Exact Inference ◽  
Modelling Studies ◽  
Approximate Bayesian ◽  
Approximate Bayesian Inference ◽  
Abc Methods

Bayesian methods are advantageous for biological modelling studies due to their ability to quantify and characterize posterior variability in model parameters. When Bayesian methods cannot be applied, due either to non-determinism in the model or limitations on system observability, approximate Bayesian computation (ABC) methods can be used to similar effect, despite producing inflated estimates of the true posterior variance. Owing to generally differing application domains, there are few studies comparing Bayesian and ABC methods, and thus there is little understanding of the properties and magnitude of this uncertainty inflation. To address this problem, we present two popular strategies for ABC sampling that we have adapted to perform exact Bayesian inference, and compare them on several model problems. We find that one sampler was impractical for exact inference due to its sensitivity to a key normalizing constant, and additionally highlight sensitivities of both samplers to various algorithmic parameters and model conditions. We conclude with a study of the O'Hara–Rudy cardiac action potential model to quantify the uncertainty amplification resulting from employing ABC using a set of clinically relevant biomarkers. We hope that this work serves to guide the implementation and comparative assessment of Bayesian and ABC sampling techniques in biological models.

Domain-Specific Language Abstractions for Compression

2021 ◽  
Author(s):  
Jessica Ray ◽  
Ajav Brahmakshatriya ◽  
Richard Wang ◽  
Shoaib Kamil ◽  
Albert Reuther ◽  
...  
Keyword(s):  
Domain Specific Language ◽  
Specific Language ◽  
Domain Specific

scholarly journals Information-Theoretic Generalization Bounds for Meta-Learning and Applications

Entropy ◽  
2021 ◽  
Vol 23 (1) ◽  
pp. 126
Author(s):  
Sharu Theresa Jose ◽  
Osvaldo Simeone
Keyword(s):  
Learning Algorithm ◽  
Broad Class ◽  
Performance Measure ◽  
Training Data ◽  
Learning To Learn ◽  
Data Set ◽  
Information Theoretic ◽  
Meta Learning ◽  
Task Training ◽  
Test Sets

Meta-learning, or “learning to learn”, refers to techniques that infer an inductive bias from data corresponding to multiple related tasks with the goal of improving the sample efficiency for new, previously unobserved, tasks. A key performance measure for meta-learning is the meta-generalization gap, that is, the difference between the average loss measured on the meta-training data and on a new, randomly selected task. This paper presents novel information-theoretic upper bounds on the meta-generalization gap. Two broad classes of meta-learning algorithms are considered that use either separate within-task training and test sets, like model agnostic meta-learning (MAML), or joint within-task training and test sets, like reptile. Extending the existing work for conventional learning, an upper bound on the meta-generalization gap is derived for the former class that depends on the mutual information (MI) between the output of the meta-learning algorithm and its input meta-training data. For the latter, the derived bound includes an additional MI between the output of the per-task learning procedure and corresponding data set to capture within-task uncertainty. Tighter bounds are then developed for the two classes via novel individual task MI (ITMI) bounds. Applications of the derived bounds are finally discussed, including a broad class of noisy iterative algorithms for meta-learning.

Sign in / Sign up

Export Citation Format

Share Document